1
Zwitterionic polymer modification of polyamide reverse-osmosis
2
membranes via surface amination and atom transfer radical
3
polymerization for anti-biofouling
4 5 6
Zhe Yang, Daisuke Saeki, Hideto Matsuyama*
7 8
Center for Membrane and Film Technology, Department of Chemical Science and
9
Engineering, Kobe University, 1-1 Rokkodai, Nada, Kobe 657-8501, Japan
10 11
*Corresponding author. Phone & FAX: +81-78-803-6180
12
E-mail:
[email protected] (D. Saeki);
[email protected] (H. Matsuyama)
1
1
Abstract
2
Surface-initiated atom transfer radical polymerization (SI-ATRP) is a powerful method to
3
uniformly modify the surface of reverse-osmosis (RO) membranes with functional polymers and
4
prevent biofouling. However, immobilization of the initiator, an essential step of SI-ATRP, is
5
difficult to perform directly on commercial polyamide RO membranes. This study describes an
6
effective pretreatment method to immobilize ATRP initiators on the surface of polyamide RO
7
membranes and the effect of the polymer chain length on the biofouling behavior. Firstly, RO
8
membrane surfaces were aminated with 3-aminopropyltrimethoxysilane (APTES). Then, α-
9
bromoisobutyryl bromide (BIBB), an acyl halide-type ATRP initiator, was reacted with the APTES
10
layer. A zwitterionic polymer, poly[(2-methacryloyloxy)ethyl]dimethyl[3-sulfopropyl]ammonium
11
hydroxide (pMEDSAH), was then grafted on the membrane surface via SI-ATRP. The APTES
12
treatment effectively improved the amount of BIBB immobilized on the membrane surface,
13
maintaining the water permeability and salt rejection properties of the RO membrane. pMEDSAH
14
grafting enhanced the surface hydrophilicity and changed the surface to a smoother and denser
15
morphology. Regarding the biofouling behavior, static bacterial adhesion on the membrane surface
16
was prevented by increasing the ATRP polymerization time. In cross-flow bacterial filtration tests,
17
the membranes grafted with pMEDSAH at polymerization times of over 1 h presented no
18
permeability decline and little biofilm coverage.
19 20
Key words: Reverse-osmosis membrane, surface-initiated atom transfer radical polymerization,
21
surface amination, zwitterionic polymer, anti-biofouling
2
1
1 Introduction
2
Polyamide reverse-osmosis (RO) membranes are on the cutting edge of membrane technology,
3
being widely applied for seawater and brackish water desalination, as well as water reuse, owing to
4
their compactness, modularity, reliability, and energy- and space-saving [1]. These membranes
5
exhibit excellent performance to purify water containing organic foulants such as bacteria, proteins,
6
and polysaccharides [2-5]. One of the major drawbacks of most polyamide RO membranes is
7
biofouling, which causes the water flux to decline, shortens the membrane lifetime, and increases
8
the energy consumption [6, 7]. Therefore, many strategies have been devised to achieve ultra-low
9
biofouling [8].
10
Increasing the hydrophilicity of membrane surfaces is regarded to be an effective method to
11
reduce biofouling because the hydration layer forms an energetic barrier to prevent foulant
12
adsorption [9-11]. Thus, extensive efforts have been devoted to investigating surface modifications
13
of polyamide RO membranes with hydrophilic materials. Poly(ethylene glycol) (PEG) has been
14
proposed as an anti-biofouling material because neutrally charged PEG grafting results in high
15
surface hydrophilicity and excluded volume effects, improving the resistance toward nonspecific
16
adsorption of organic foulants [12]. However, PEG tends to auto-oxidate in the presence of oxygen,
17
losing its antifouling ability [13].
18
Over recent years, zwitterionic polymers such as polyphosphobetaine, polysulfobetaine, and
19
polycarboxybetaine have been considered as excellent anti-biofouling materials [14-17].
20
Zwitterionic species containing both positively and negatively charged groups are able to bind water
21
molecules even more strongly and stably than other hydrophilic materials via electrostatically
22
induced hydration [18]. On zwitterionic polymer-grafted surfaces, the high water content prevents
23
the irreversible adsorption of organic foulants without significant conformational changes [19].
24
Therefore, many researches about the surface modification of RO membranes using zwitterionic
25
polymers were conducted for anti-biofouling. Wang et al. modified the RO membrane surface with
26
polycarboxybetaine via redox-initiated reaction, and reported that the permselectivity, anti-
27
biofouling and cleaning properties of RO membranes were significantly improved [20]. Gleason et
28
al. reported the RO membranes modified with polycarboxybetaine [21] or polysulfobetaine [22] via
3
1
initiated chemical vapor deposition technique effectively prevented the bacterial adhesion. Azari et
2
al. incorporated a redox functional amino acid 3-(3,4-dihydroxyphenyl)-L-alanine onto commercial
3
RO membranes to create a zwitterionic surface that resists membrane fouling [23]. As above, the
4
surface modification using zwitterionic polymer is a promising way to prevent the bacterial
5
adhesion. However, the effect of the polymer grafting and polymer chain length on the dynamic
6
biofouling behavior for water filtration applications has still not been clarified. The difficulty to
7
evaluate the anti-biofouling properties is time-dependent, slow and complicated behavior of
8
biofouling [24]. Biofouling is sequential phenomenon corresponding to bacterial adhesion, growth,
9
biofilm formation, and pore blocking. In the previous studies, the protein-containing water [20, 23],
10
bacterial suspended water [25-28] or actual waste water like as membrane bioreactor-treated water
11
[29, 30] were used as the feed water. These systems wouldn’t reflect the biofouling behavior
12
including bacterial adhesion, bacterial growth and biofilm formation on the membrane surface.
13
There are several preparation approaches for commercial polyamide RO membrane with
14
zwitterionic polymers, including redox reactions [20, 31], electrostatic coating [32], UV-initiated
15
radical grafting [33, 34], and initiated chemical vapor deposition [21, 22, 35]. However, these
16
modification methods have some major disadvantages, such as low surface density, low stability,
17
and degradation of the separation performance. Notably, all these approaches are limited to increase
18
the grafted-polymer density and control the polymer chain length precisely. In contrast, surface-
19
initiated atom transfer radical polymerization (SI-ATRP) is a promising method to accurately and
20
efficiently control the chain length of grafted polymers on material surfaces [36]. Recently, SI-
21
ATRP has been applied for the surface modification of water purification membranes for specific
22
applications, such as antifouling, stimulus-responsiveness, adsorption functionalities, and
23
pervaporation [37]. Initiator immobilization on the membrane surface is a crucial step of SI-ATRP
24
reactions. The acyl halide-type initiator α-bromoisobutyryl bromide (BIBB) is generally used
25
because acyl halide groups react readily with the hydroxyl and amino groups of membrane surfaces.
26
However, commercial polyamide RO membranes have a small number of these terminal groups on
27
their surface and are less reactive, limiting the reaction with such initiators. To apply SI-ATRP to
28
polyamide RO membranes, the modification of the membrane fabrication process is necessary [25,
4
1
38, 39] and is less acceptable to the commercialization. Thus, additional efforts are needed to
2
improve the immobilization of ATRP initiators for easy application on the polyamide RO membrane.
3
Our objectives in this work are firstly to develop an effective method to immobilize SI-ATRP
4
initiators on the less reactive surface of polyamide RO membranes, and secondly to evaluate the
5
effect of the chain length of a zwitterionic polymer on biofouling prevention systematically. 3-
6
Aminopropyltrimethoxysilane (APTES) was used to aminate the surface of a polyamide RO
7
membrane to improve the BIBB immobilization. Then, after BIBB immobilization, pMEDSAH was
8
grafted on the membrane surface via SI-ATRP. The chain length of the grafted zwitterionic polymer
9
was controlled by changing polymerization time and quantified by gel permeation chromatography
10
(GPC). The surface chemical properties of the modified membranes with various polymer chain
11
lengths were characterized by attenuated total reflection-Fourier transform infrared (ATR-FTIR)
12
spectroscopy, X-ray photoelectron spectroscopy (XPS), and contact angle measurements. The
13
surface morphology was observed using scanning electron microscopy (FE-SEM) and atomic force
14
microscopy (AFM). The biofouling resistance of the modified membranes was evaluated by static
15
bacterial adhesion and dynamic biofouling filtration experiments.
16 17
2 Experimental
18
2.1 Materials
19
Commercial polyamide RO membranes (ES20) were purchased from Nitto Denko (Osaka, Japan).
20
3-Aminopropyltrimethoxysilane (APTES) was used to modify the surface of the RO membranes.
21
α-Bromoisobutyryl bromide (BIBB; Sigma-Aldrich, St. Louis, MO, USA) was used as the initiator
22
for SI-ATRP. L-Ascorbic acid (Tokyo Chemical Industry, Tokyo, Japan), copper (II) bromide
23
(CuBr2), and tris(2-pyridylmethyl)amine (TPMA) were used for the ATRP reaction. MEDSAH
24
(Sigma-Aldrich) was used as the zwitterionic monomer. Ethyl-2-bromoisobutyrate (EBIB; Tokyo
25
Chemical Industry) was used as the ATRP initiator for polymerization in solution. Sphingomonas
26
paucimobilis NBRC 13935 was obtained from the NITE Biological Resource Center (Chiba, Japan)
27
and used as the model bacteria strain. Tryptic soy broth (TSB; Becton, Dickinson and Company,
28
Franklin Lakes, NJ, USA) was used to culture the bacteria and SYTO9 (Life Technologies, Carlsbad,
5
1
CA, USA) to stain them. All other chemicals were purchased from Wako Pure Chemical Industry,
2
Osaka, Japan. Deionized (DI) water was obtained from a Milli-Q water purification system (Merck
3
Millipore, Darmstadt, Germany) and used in all the experiments.
4 5
2.2 Membrane modification
6
A scheme of the membrane modification process is presented in Fig. 1. First, a pretreatment was
7
performed to facilitate the BIBB immobilization. A circular polyamide RO membrane with 36-mm
8
diameter was immersed in an aqueous solution of APTES and then subjected to vacuum at room
9
temperature for 10 min. Then, the membrane was immersed in a hexane solution of BIBB at room
10
temperature for 1 min and rinsed thoroughly with hexane and DI water.
11
For the SI-ATRP reaction, the BIBB-immobilized membrane was placed in a glass bottle. A
12
mixture of 14.0 mL DI water and methanol (1:1, v/v) containing 10 mmol of MEDSAH and 0.8
13
mmol of ascorbic acid were added to the bottle. After 10 min of nitrogen bubbling, 4 mL of DI
14
water/methanol (1:1, v/v) containing 0.02 mmol of CuBr2 and 0.04 mmol of TPMA was added to
15
initiate the ATRP reaction. The mixed solution was then stirred for a given polymerization time,
16
after which an RO membranes grafted with MEDSAH polymer (pMEDSAH) was obtained. The
17
membrane was washed with DI water in a shaker overnight at 40 ℃ and kept in DI water for further
18
use.
19 20
The BIBB-immobilized membranes are denoted by xSyB, in which x and y indicate the concentration of APTES (vol%) and BIBB (wt%) during membrane fabrication, respectively. Polyamide membrane
APTES + H2O
BIBB
MEDSAH
Br Br
Br Br
Si-ATRP
21
10 min
Br
Br
Br
1 min
22
Fig. 1. Scheme of the SI-ATRP reaction of MEDSAH on a polyamide RO membrane after APTES
23
pretreatment.
6
1 2
2.3 Surface characterization
3
2.3.1 Surface morphology
4
The surface morphology of the fabricated membranes was observed by both field emission
5
scanning electron microscopy (FE-SEM; JSF-7500, JEOL, Tokyo, Japan) and atomic force
6
microscopy (AFM; SPA-400, Hitachi High-Technologies, Tokyo, Japan). The surface roughness Ra
7
was determined by AFM using the dynamic force mode and a SI-DF40 cantilever. For FE-SEM and
8
AFM observation, the membrane samples were freeze-dried under vacuum using a freeze-dryer
9
(FDU-1200, Tokyo Rikakikai, Tokyo, Japan).
10 11
2.3.2 Chemical and physical properties
12
The surface chemistry of the pristine and modified membranes was evaluated by ATR-FTIR
13
spectroscopy (Nicolet iS5, Thermo Fisher Scientific, MA, USA). The surface elemental content of
14
the membranes was evaluated by XPS (JPS-9010MC, JEOL). Before the measurements, the
15
membranes were completely dried. The surface hydrophilicity of the membranes was determined
16
by the water droplet method in a contact angle meter (DM-300, Kyowa Interface Science, Saitama,
17
Japan). The reported contact angle values are the average of at least two different membrane samples
18
after ten measurements of each sample.
19 20
2.4 Quantification of grafted pMEDSAH
21
The chain length of the pMEDSAH grafted on the membrane surface was evaluated from ATRP
22
in solution using EBIB as the ATRP initiator. The average molecular weight of the synthesized
23
pMEDSAH was determined by a GPC apparatus (Viscotek TDAmax, Malvern Instruments,
24
Worcestershire, UK) and a Shodex GF-510HQ column (Showa Denko, Tokyo, Japan).
25 26
2.5 Membrane performance
27
2.5.1 Water permeability and salt rejection
7
1
The water permeability and salt rejection properties of the fabricated membranes were evaluated
2
using a laboratory scale cross-flow membrane test unit (Fig. 2), as described in our previous study
3
[40]. The effective surface area of the membrane was 8.0 cm2. DI water and aqueous 0.05 wt%
4
NaCl were used as the feed water to measure the water permeability and salt rejection, respectively.
5
The feed water was introduced at 2.0 mL/min using a plunger pump (NPL-120, Nihon Seimitsu
6
Kagaku, Tokyo, Japan). The applied pressure on the membrane was controlled using a back-
7
pressure valve at 0.75 MPa. The feed water side of the membrane surface was magnetically stirred
8
at 500 rpm. The water permeability (L; L/(m2 h MPa); LMH/MPa) was calculated by:
9
𝐿𝐿 =
𝑄𝑄
(1)
𝐴𝐴×𝑡𝑡×𝑃𝑃
10
where Q is the volume of accumulated permeate, A is the effective surface area of the membrane, t
11
is the time, and P is the applied pressure on the membrane surface. In the case of salt rejection
12
measurements, the electric conductivity of the feed (Cf) and permeate (Cp) water was monitored
13
with a conductivity meter (B-771, Horiba, Kyoto, Japan) and the salt rejection (R) was calculated
14
as follows:
15
𝑅𝑅 (%) =
𝐶𝐶f −𝐶𝐶p 𝐶𝐶f
× 100
(2)
Permeate
PC
MPa
Feed water
Pump
16
Pressure gauge
Magnetic stirrer
Valve
Condensate
17
Fig. 2. Scheme of the laboratory-scale cross-flow membrane test unit to evaluate the membrane
18
performance.
19 20
2.5.2 Static bacteria adhesion tests
21
The bacterial adhesion propensity of the fabricated membranes was evaluated by static bacteria
22
adhesion tests, as described in our previous study [32]. Bacteria were precultured in TSB medium 8
1
for 12 h at 30 ℃. Then, the precultured bacterial suspension was diluted 50 times with fresh TSB
2
medium and cultured again for 4 h at 30 ℃. The bacterial suspension was diluted with fresh TSB
3
medium to a final optical density of 0.05 at 450 nm for each of the experiments. The optical density
4
was measured using a spectrophotometer (V-650, Jasco, Tokyo, Japan). The membranes were
5
immersed in a bacteria suspension at 120 rpm at 30 ℃ for 24 h. The membranes were gently rinsed
6
twice with aqueous 0.85 wt% NaCl and immersed in aqueous 0.85 wt% NaCl containing SYTO9
7
for 20 min to stain the bacteria adhered to the membrane surface. To fix the stained bacteria on the
8
membrane surface, the membranes were immersed in aqueous 2.5% glutaraldehyde for 3 min, and
9
rinsed and kept in aqueous 0.85 wt% NaCl. The stained membranes were observed using a confocal
10
laser scanning microscopy (CLSM; FV1000D, Olympus, Tokyo, Japan) and the coverage of
11
adhered bacteria was calculated with the Image J software (National Institutes of Health, Bethesda,
12
MD, USA).
13 14
2.5.3 Dynamic biofouling filtration tests
15
The biofouling propensity of pristine and modified membranes was assessed by dynamic
16
biofouling tests, as described in our previous study [41]. This evaluation system can simulate the
17
biofouling behavior of bacterial growth, biofilm formation, and flux decline in a short experimental
18
period. A bacterial suspension was prepared similarly to that for the static bacterial adhesion tests.
19
First, bacterial solution was poured on the surface of the fabricated membrane and incubated at 30 ℃
20
for 1 h for adhesion of bacteria on the membrane surface. The membrane was briefly washed with
21
aqueous 0.85 wt% NaCl and set in a cross-flow membrane test unit (Fig. 2). TSB medium, diluted
22
five times with aqueous 0.85 wt% NaCl, was fed to the membrane cell at 2.0 mL/min and 1.5 MPa
23
for 20 h. The organic concentration of this feed water is higher than that of actual waste water to
24
accelerate the biofouling [42, 43]. The feed water side of the membrane surface was magnetically
25
stirred at 200 rpm and the whole membrane cell was incubated at 30 ℃. The accumulated permeate
26
weight was recorded to calculate the changes in the permeability. After the filtration experiment,
27
the tested membrane was washed with aqueous 0.85 wt% NaCl and the adhered bacteria were
28
stained with SYTO9, as described for the static bacterial adhesion tests. The stained membrane
9
1
surfaces were observed using CLSM and FE-SEM. The bacterial coverage was calculated with the
2
Image J software.
3 4
3 Results and Discussion
5
3.1 Effect of APTES treatment on BIBB immobilization
6
In order to improve the initiator immobilization on the polyamide RO membrane, different
7
pretreatments were studied. Table 1 shows the effect of the pretreatments on the membrane
8
performance and Br ratio (measured by XPS) on the membrane surface after BIBB immobilization.
9
The BIBB-modified membrane without any pretreatment exhibited a similar membrane
10
performance and Br ratio than the pristine membrane. BIBB is easily decomposed by water and
11
loses reactivity; therefore, BIBB was barely immobilized on the surface of the membrane covered
12
with water molecules. Drying the membrane before BIBB immobilization was also attempted as a
13
pretreatment. Although the BIBB content increased compared to that of the membrane without
14
pretreatment, both the water permeability and salt rejection decreased, indicating that the membrane
15
structure collapsed upon drying. When the APTES-treated membrane was reacted with BIBB, the
16
Br ratio increased while the membrane performance was maintained. The APTES treatment affords
17
a polysiloxane layer on the membrane surface via a sol–gel reaction that improves the salt rejection
18
[44], while simultaneously introducing amino groups able to react easily with BIBB. Therefore, the
19
APTES treatment was applied as a SI-ATRP pretreatment throughout the work described here.
20 21
Table 1. Effect of pretreatments before BIBB immobilization on the water permeability, salt
22
rejection, and Br ratio of the fabricated membranes Permeability
NaCl
Br 3d5/2
Pretreatment (LMH/MPa)
rejection (%)
(atomic %)**
Pristine membrane*
67.48
97.3
0
No pretreatment
49.37
98.3
0.04
Drying
36.82
83.0
2.04
1 wt% APTES treatment
50.48
96.9
1.98
10
All the results are the average of three measurements with less than 5% variation. The BIBB concentration was 3 wt%. * Without BIBB immobilization. ** Elemental ratio measured by XPS.
1 2
3.2 Effect of APTES and BIBB concentration on BIBB immobilization
3
In order to optimize the conditions for BIBB immobilization, the effect of the concentration of
4
APTES and BIBB on the water permeability, salt rejection, and surface elemental ratio was
5
investigated. The water permeability and salt rejection properties of BIBB-immobilized membranes
6
at different APTES concentrations are presented in Fig. 3. With the increasing APTES concentration,
7
the water permeability decreased, while there were no significant changes in the salt rejection. High
8
APTES concentrations reduce the surface hydrophilicity due to the larger number of carbon groups,
9
affording a reduction of the water permeability and an increase of the salt rejection [44]. Table 2
10
shows the elemental ratios of BIBB-immobilized membranes at different APTES concentrations.
11
The Si and Br ratios increased with the APTES concentration, corresponding to the introduction of
12
APTES and BIBB on the membrane surface, respectively. 100 Water permeability
Salt rejection
60
95
40
90
20
85
0
80 RO
13
Salt rejection (%)
Water permeability (LMH/MPa)
80
0.5S3B
1S3B
2S3B
3S3B
14
Fig. 3. Effect of the APTES concentration on the water permeability and salt rejection of pristine
15
and BIBB-immobilized membranes. The BIBB concentration was fixed at 3.0 wt%.
16 17
Table 2. Effect of the APTES concentration on the surface elemental ratios of pristine and BIBB-
18
immobilized membranes
11
Elemental ratio (atomic %) Membrane
C 1s
O 1s
N 1s
Si 2p3/2
Br 3d5/2
Pristine membrane
70.74
16.42
12.84
0.00
0.00
0.5S3B
69.79
15.97
10.95
2.17
0.98
1S3B
67.67
17.01
10.57
2.77
1.98
2S3B
64.72
17.75
11.16
3.68
2.70
3S3B
65.64
16.95
10.17
4.09
3.15
1 2
Next, the effect of the BIBB concentration on the BIBB immobilization was investigated. Fig. 4
3
shows the performance of membranes fabricated at different BIBB concentrations. The salt rejection
4
barely changed with the increasing BIBB concentration, although the water permeability of the
5
BIBB-immobilized membranes slightly decreased compared to that of the pristine membrane. The
6
BIBB layer hinders the permeation of water molecules [25]. The elemental ratios of the membrane
7
surfaces are presented in Table 3, which shows how the Br ratio increases with the increasing BIBB
8
concentration, indicating that the APTES treatment promotes the introduction of amino groups, thus
9
facilitating the immobilization of BIBB on the surface of the polyamide RO membranes.
10
In the following, BIBB immobilization was carried out with 1 vol% of APTES and 3 wt% of
11
BIBB, as these conditions resulted in the optimal immobilization of BIBB on the surface of the
12
commercial polyamide RO membrane while maintaining its performance at satisfactory levels. 100 Water permeability
Salt rejection
60
95
40
90
20
85
0
13
Salt rejection (%)
Water permeability (LMH/MPa)
80
80 RO
1S1B
1S2B
1S3B
1S5B
14
Fig. 4. Effect of the BIBB concentration on the water permeability and salt rejection of pristine and
15
BIBB-immobilized membranes. The APTES concentration was fixed at 1.0 vol%.
16 12
1
Table 3. Effect of the BIBB concentration on the elemental ratios of pristine and BIBB-immobilized
2
membranes Membrane
Elemental ratio (atomic %) C 1s
O 1s
N 1s
Si 2p3/2
Br 3d5/2
70.74
16.42
12.84
0.00
0.00
1S1B
63.97
19.05
11.24
2.12
1.01
1S2B
69.87
15.57
11.02
2.54
1.49
1S3B
67.67
17.01
10.57
2.77
1.98
1S5B
68.22
16.81
10.23
2.22
2.52
Pristine membrane
3 4
3.3 Effect of the polymerization time on the SI-ATRP of the RO membrane surface
5
We investigated the effect of the polymerization time of SI-ATRP on the structure of the RO
6
membranes. Figures 5 and 6 show the surface morphology and average roughness of the membrane
7
surfaces, respectively. The pristine and BIBB-immobilized membranes presented ridge-and-valley
8
structures. The surface roughness increased significantly upon BIBB immobilization probably due
9
to the formation of APTES aggregates during the immobilization process. With the increasing
10
polymerization time, the surface structure became smoother and the surface roughness decreased.
11
It seems that pMEDSAH with longer chain lengths is able to fill the ridges and valleys of the
12
polyamide layer [45]. pMEDSAH-grafted membranes Pristine RO
1S3B 20 min
30 min
60 min
120 min
AFM
FE-SEM
10 min
13 14
Fig. 5. FE-SEM and AFM images of the pristine (RO), BIBB-immobilized (1S3B), and
15
pMEDSAH-grafted membranes fabricated at different polymerization times. The scale bar in the
16
SEM images indicates 500 nm. 13
Average roughness (nm)
160 120 80 40 0
1 2
RO 1S3B 10 20 30 60 Polymerization time (min)
120
Fig. 6. Surface roughness of the fabricated membranes determined by AFM.
3 4
Fig. 7(a) shows the ATR-FTIR spectra of the membranes subjected to different polymerization
5
times. The chemical composition of the membrane surface was not significantly changed upon
6
BIBB immobilization, as reported in a previous study [25]. The characteristic bands of the main
7
functional groups of MEDSAH (Fig. 7(b)) are observed in the spectrum of the membrane after 10
8
min of ATRP at 1720, 1039, and 953 cm-1, attributed to carbonyl, sulfonate, and quaternary amine
9
groups, respectively [46, 47]. The absorbance changes in these peaks are presented in Fig. 7(c). The
10
intensity of all these bands increased with the polymerization time. In particular, the absorbance of
11
the sulfonate group at 1039 cm-1 increased more dramatically than that of the other functional groups.
12
Furthermore, the element sulfur was only detected by XPS in the pMEDSAH-grafted membranes
13
(Fig. 8), whose ratio increased with the polymerization time. These results indicate that the chain
14
length of pMEDSAH on the membrane surface increases with the polymerization time. We
15
estimated the weight-average molecular weight of pMEDSAH on the membrane surface by GPC.
16
The ATRP reaction was carried out in solution using EIBB as the ATRP initiator. The weight-
17
average molecular weight of pMEDSAH gradually increased from 22054 to 60811 kDa upon
18
elongating the polymerization time from 30 to 120 min, indicating that the chain length of
19
pMEDSAH also increased with the polymerization time.
14
1 2
Fig. 7. (a) ATR-FTIR spectra of pristine (RO), BIBB-immobilized (1S3B), and pMEDSAH-grafted
3
membranes fabricated at different polymerization times; (b) chemical structure of pMEDSAH and
4
the absorbance wavelengths of its functional groups; and (c) variation in the absorbance of the
5
different functional groups with the polymerization time.
Sulfur composition (%)
4
3
2
1 0
6 7
30 60 90 Polymerization time (min)
120
Fig. 8. Sulfur content in pMEDSAH-grafted membranes at different polymerization times.
8 9
Water contact angle measurements were carried out to investigate the surface hydrophilicity of
10
the fabricated membranes. Fig. 9 shows the water contact angle of the fabricated membranes. The
11
water contact angle significantly decreased after BIBB immobilization (1S3B) and ATRP treatment.
12
The short hydrocarbon chain of APTES reduced the surface hydrophobicity [44]. Upon pMEDSAH
13
grafting, the water contact angle was further reduced and the membrane surface became hydrophilic
14
due to the strong hydration capacity of the zwitterionic polymer [48]. The hydrophilicity of the
15
membrane surface slightly increased with the increasing polymerization time. These results further
15
1
confirm the introduction of the zwitterionic pMEDSAH on the surface of polyamine RO membranes
2
is very useful to make the membrane surface more hydrophilic.
3 4
Fig. 9. Contact angle of pristine RO and pMEDSAH-grafted membranes at increasing
5
polymerization times.
6 7
The water permeability and salt rejection properties of the fabricated membranes are presented
8
in Fig. 10. With the increasing polymerization time, the water permeability decreased gradually
9
owing to a thick layer of pMEDSAH formed on the membrane surface, resulting in hindered water
10
permeation. On the other hand, the salt rejection barely changed upon pMEDSAH grafting.
11 12
Fig. 10. Water permeability and salt rejection properties of pMEDSAH-grafted membranes at
13
different polymerization times.
14 15
3.4 Static bacterial adhesion on pMEDSAH-modified membranes
16
Resistance to microorganism adhesion is critical for biofouling mitigation since the adhesion of
17
microorganisms typically leads to subsequent colonization and formation of biofilms [49]. The
18
bacterial adhesion data for the pristine and pMEDSAH-grafted membranes at different
16
1
polymerization times from static adhesion tests are shown in Fig. 11. The degree of bacterial
2
adhesion was clearly prevented by pMEDSAH-grafting, which decreased with the increasing
3
polymerization time, similarly to the case of the surface hydrophilicity. The pMEDSAH-grafted
4
membranes also effectively prevented the protein adsorption (Fig. S1) as same as previously
5
reported [38, 50]. The tendency of the prevention of the protein adsorption showed a same manner
6
as that of the bacterial adhesion, suggesting that the bacterial adhesion is partially caused by the
7
proteins on their surface [51]. As the membrane surface becomes more hydrophilic, a hydration
8
layer is easily formed on the membrane surface, which prevents the adsorption and deposition of
9
hydrophobic bacteria on the membrane surface, thus reducing fouling [52].
Bacteria adhesion (%)
12 RO 1S3B
8
4
0 0
30 60 90 Polymerization time (min)
120
10 11
Fig. 11. Bacterial adhesion on pristine and pMEDSAH- grafted membranes at different
12
polymerization times.
13 14
3.5 Dynamic biofouling
15
Finally, we investigated the effect of the chain length of the grafted pMEDSAH on the biofouling
16
behavior of RO membranes. Fig. 12 shows the changes in the water permeability from cross-flow
17
bacterial filtration tests. The permeability of the pristine membrane remarkably decreased after 10
18
h, suggesting the biofilm formation and pore blocking. On the other hand, that of the pMEDSAH-
19
grafted membranes, especially with over 60-min polymerization, didn’t showed the permeability
20
decline at 10 h. After 20-h filtration experiments, the water permeability of the pMEDSAH-grafted
21
membranes was higher than that of the pristine membrane, which decreased to values below 40%
22
of the initial value. Such a permeability reduction was inhibited upon pMEDSAH grafting, and the
23
suppression degree increased with the increasing polymerization time. At polymerization times of 17
1
10, 30, and 40 min, the water permeability decreased slightly after 20 h of filtration. On the other
2
hand, when the polymerization time was over 60 min, the pMEDSAH-grafted membranes
3
maintained the initial water permeability throughout the filtration test. Although the membranes
4
modified at 60 min of polymerization time present a lower initial permeability than the pristine
5
membrane due to the surface modification, it exhibited the highest water permeability after 20 h of
6
filtration because of the completely prevented biofouling. Water permeability (LHM/Mpa)
20
15
10
5 RO 40 min
10 min 60 min
30 min 120 min
0 0
7
5
10 15 Filtration time (h)
20
8
Fig. 12. Time course for the water flux of pristine and pMEDSAH-grafted membranes from
9
biofouling filtration tests using a cross-flow membrane test unit.
10 11
Fig. 13 shows the biofilm structure on the tested membrane surfaces obtained by confocal
12
microscopy and SEM. The surface of the pristine membrane appears significantly covered with
13
bacteria, indicating that bacteria easily adhere to the pristine membrane surface forming a biofilm.
14
With the increasing polymerization time, the biofilm becomes smaller and thinner, as well as less
15
dense. Fig. 14 compiles the bacterial adhesion data obtained from static adhesion tests and dynamic
16
biofouling tests. The bacterial coverage from both static bacterial adhesion and dynamic biofouling
17
tests decreased with the increasing polymerization time. However, the relationship between two
18
correlated data was not a linear correlation. Regarding the pMEDSAH-grafted membrane after 1 h
19
of polymerization, the bacterial adhesion in the dynamic biofouling test was prevented more
20
significantly than that in the static adhesion test. The chain length of the grafted polymer is one of
21
the important factors influencing the biofouling behavior. pMEDSAH with short chain length
22
reduces the adhesion force of bacteria on the polyamide RO membrane surface and prevents the 18
1
bacterial adhesion, although it barely reduced the biofilm growth in the dynamic tests, while
2
pMEDSAH with longer chain lengths is able to prevent both bacterial adhesion and biofilm growth. pMEDSAH-grafted membranes 10 min
30 min
40 min
60 min
120 min
3
(b) FE-SEM
(a) CLSM
Pristine RO
4
Fig. 13. Bacteria adhesion on the surface of pMEDSAH-grafted membranes at different
5
polymerization times after dynamic biofouling filtration tests. 3D images obtained by (a) CLSM
6
and (b) FE-SEM. The scale bar in the FE-SEM images indicates 50 μm. 1 Normalized bacterial coverage in dynamic biofouling test [-]
10 min
Pristine membrane
0.8 30 min
0.6
0.4
0.2 120 min 60 min 0 0
7
0.2 0.4 0.6 0.8 Normalized bacterial coverage in static adhesion test [-]
1
8
Fig. 14. Relationship between the bacterial adhesion from static bacterial adhesion tests and
9
dynamic biofouling filtration tests. The bacterial coverage on the pMEDSAH-grafted membranes
10
was normalized to that on the pristine membrane. The bacterial coverage in the dynamic biofouling
11
tests was obtained by analyzing the CLSM images with the Image J software.
12 13
The poly-zwitterionic materials used in this study have both positively and negatively charged
14
moieties on the same monomer unit, resulting in uniform charge distribution and neutrality on the
19
1
membrane surface. Through electrostatically induced hydration, zwitterionic materials are able to
2
bind water molecules even more strongly and stably than other hydrophilic materials such as PEG
3
and polyvinyl alcohol. In this way, poly-zwitterionic materials with suitable chain lengths are able
4
to maximize the surface hydration and reduce the electrostatic interactions with the foulant [18].
5
Moreover, the dense and smooth coverage obtained with the zwitterionic polymer with longer chain
6
lengths also leads to excellent biofouling resistance owing to the lower surface roughness, which
7
reduces the surface area for membrane–foulant interactions [53, 54]. In addition, the movement of
8
pMEDSAH chain will be easier and larger in the case of longer chain [55], which is another reason
9
for the prevention of biofouling. Thus, control of the architecture of the grafted polymer is crucial
10
to develop anti-biofouling membranes.
11 12
4 Conclusions
13
In this work, a surface modification method for polyamide RO membranes with a zwitterionic
14
polymer via aminosilane treatment and SI-ATRP was proposed for surface amination and anti-
15
biofouling purposes. The APTES treatment effectively introduced amino groups on the polyamide
16
layer of the RO membrane and facilitated the immobilization of an ATRP initiator (BIBB), while
17
maintaining the water flux and salt rejection properties of the original membrane. A zwitterionic
18
polymer, pMEDSAH, was grafted on the surface of BIBB-immobilized membranes and its chain
19
length was found to increase with the polymerization time. The hydrophilicity and static bacterial
20
adhesion of the membranes was improved by increasing the polymerization time. The dynamic
21
biofouling tests showed that biofilm formation and biofouling were effectively prevented by the
22
pMEDSAH grafting and the increase of polymerization time.
23
Our strategy using APTES treatment is applicable for facilitating the ATRP initiator
24
immobilization on less reactive surfaces of commercial RO membranes maintaining the permeation
25
properties. Furthermore, this study not only demonstrated the great potential of the zwitterionic
26
polymer modification of RO membranes for the biofouling prevention, but also clarified that the
27
control of the pMEDSAH chain length is necessary to obtain the best performance for the
28
improvement of the anti-biofouling property.
20
1 2
Acknowledgements
3
This work was supported in part by a Research Grant from Kurita Water and Environment
4
Foundation (No. 16A073) to D.S., and Special Coordination Fund for Promoting Science and
5
Technology, Creation of Innovation Centers for Advanced Interdisciplinary Research Areas
6
(Innovative Bioproduction, Kobe) from the Ministry of Education, Culture, Sports, Science and
7
Technology of Japan.
8 9
References
10
[1] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology, and the
11
environment, Science, 333 (2011) 712-717.
12
[2] C. Leroy, C. Delbarre, F. Ghillebaert, C. Compere, D. Combes, Influence of subtilisin on the
13
adhesion of a marine bacterium which produces mainly proteins as extracellular polymers, J. Appl.
14
Microbiol., 105 (2008) 791-799.
15
[3] S. Lee, M. Elimelech, Relating organic fouling of reverse osmosis membranes to intermolecular
16
adhesion forces, Environ. Sci. Technol., 40 (2006) 980-987.
17
[4] M. Herzberg, M. Elimelech, Biofouling of reverse osmosis membranes: role of biofilm-
18
enhanced osmotic pressure, J. Membr. Sci., 295 (2007) 11-20.
19
[5] W.S. Ang, A. Tiraferri, K.L. Chen, M. Elimelech, Fouling and cleaning of RO membranes
20
fouled by mixtures of organic foulants simulating wastewater effluent, J. Membr. Sci., 376 (2011)
21
196-206.
22
[6] Y. Mo, A. Tiraferri, N.Y. Yip, A. Adout, X. Huang, M. Elimelech, Improved antifouling
23
properties of polyamide nanofiltration membranes by reducing the density of surface carboxyl
24
groups, Environ. Sci. Technol., 46 (2012) 13253-13261.
25
[7] W. Guo, H.-H. Ngo, J. Li, A mini-review on membrane fouling, Bioresour. Technol., 122 (2012)
26
27-34.
21
1
[8] G.-R. Xu, J.-N. Wang, C.-J. Li, Strategies for improving the performance of the polyamide thin
2
film composite (PA-TFC) reverse osmosis (RO) membranes: Surface modifications and
3
nanoparticles incorporations, Desalination, 328 (2013) 83-100.
4
[9] E. Ostuni, R.G. Chapman, R.E. Holmlin, S. Takayama, G.M. Whitesides, A survey of structure−
5
property relationships of surfaces that resist the adsorption of protein, Langmuir, 17 (2001) 5605-
6
5620.
7
[10] M. He, K. Gao, L. Zhou, Z. Jiao, M. Wu, J. Cao, X. You, Z. Cai, Y. Su, Z. Jiang, Zwitterionic
8
materials for antifouling membrane surface construction, Acta Biomater., 40 (2016) 142-152.
9
[11] W. Chen, Y. Su, J. Peng, X. Zhao, Z. Jiang, Y. Dong, Y. Zhang, Y. Liang, J. Liu, Efficient
10
wastewater treatment by membranes through constructing tunable antifouling membrane surfaces,
11
Environ. Sci. Technol., 45 (2011) 6545-6552.
12
[12] G. Kang, M. Liu, B. Lin, Y. Cao, Q. Yuan, A novel method of surface modification on thin-
13
film composite reverse osmosis membrane by grafting poly (ethylene glycol), Polymer, 48 (2007)
14
1165-1170.
15
[13] I. Banerjee, R.C. Pangule, R.S. Kane, Antifouling coatings: recent developments in the design
16
of surfaces that prevent fouling by proteins, bacteria, and marine organisms, Adv. Mater., 23 (2011)
17
690-718.
18
[14] Y. Chang, S. Chen, Q. Yu, Z. Zhang, M. Bernards, S. Jiang, Development of biocompatible
19
interpenetrating polymer networks containing a sulfobetaine-based polymer and a segmented
20
polyurethane for protein resistance, Biomacromolecules, 8 (2007) 122-127.
21
[15] S. Chen, F. Yu, Q. Yu, Y. He, S. Jiang, Strong resistance of a thin crystalline layer of balanced
22
charged groups to protein adsorption, Langmuir, 22 (2006) 8186-8191.
23
[16] R.E. Holmlin, X. Chen, R.G. Chapman, S. Takayama, G.M. Whitesides, Zwitterionic SAMs
24
that resist nonspecific adsorption of protein from aqueous buffer, Langmuir, 17 (2001) 2841-2850.
25
[17] Z. Zhang, S. Chen, Y. Chang, S. Jiang, Surface grafted sulfobetaine polymers via atom transfer
26
radical polymerization as superlow fouling coatings, J. Phys. Chem. B, 110 (2006) 10799-10804.
27
[18] S. Chen, L. Li, C. Zhao, J. Zheng, Surface hydration: principles and applications toward low-
28
fouling/nonfouling biomaterials, Polymer, 51 (2010) 5283-5293.
22
1
[19] K. Ishihara, H. Nomura, T. Mihara, K. Kurita, Y. Iwasaki, N. Nakabayashi, Why do
2
phospholipid polymers reduce protein adsorption?, J. Biomed. Mater. Res., 39 (1998) 323-330.
3
[20] J. Wang, Z. Wang, J. Wang, S. Wang, Improving the water flux and bio-fouling resistance of
4
reverse osmosis (RO) membrane through surface modification by zwitterionic polymer, J. Membr.
5
Sci., 493 (2015) 188-199.
6
[21] H.Z. Shafi, A. Matin, Z. Khan, A. Khalil, K.K. Gleason, Surface modification of reverse
7
osmosis membranes with zwitterionic coatings: A potential strategy for control of biofouling, Surf.
8
Coat. Technol., 279 (2015) 171-179.
9
[22] R. Yang, J. Xu, G. Ozaydin-Ince, S.Y. Wong, K.K. Gleason, Surface-Tethered Zwitterionic
10
Ultrathin Antifouling Coatings on Reverse Osmosis Membranes by Initiated Chemical Vapor
11
Deposition, Chem. Mater., 23 (2011) 1263-1272.
12
[23] S. Azari, L. Zou, Fouling resistant zwitterionic surface modification of reverse osmosis
13
membranes using amino acid l-cysteine, Desalination, 324 (2013) 79-86.
14
[24] H.-C. Flemming, Reverse osmosis membrane biofouling, Exp. Therm Fluid Sci., 14 (1997)
15
382-391.
16
[25] D. Saeki, T. Tanimoto, H. Matsuyama, Anti-biofouling of polyamide reverse osmosis
17
membranes using phosphorylcholine polymer grafted by surface-initiated atom transfer radical
18
polymerization, Desalination, 350 (2014) 21-27.
19
[26] Y. Baek, J. Yu, S.-H. Kim, S. Lee, J. Yoon, Effect of surface properties of reverse osmosis
20
membranes on biofouling occurrence under filtration conditions, J. Membr. Sci., 382 (2011) 91-99.
21
[27] X. Zhu, H.-E. Loo, R. Bai, A novel membrane showing both hydrophilic and oleophobic
22
surface properties and its non-fouling performances for potential water treatment applications, J.
23
Membr. Sci., 436 (2013) 47-56.
24
[28] F. Razi, I. Sawada, Y. Ohmukai, T. Maruyama, H. Matsuyama, The improvement of
25
antibiofouling efficiency of polyethersulfone membrane by functionalization with zwitterionic
26
monomers, J. Membr. Sci., 401 (2012) 292-299.
23
1
[29] J. Lee, H.-R. Chae, Y.J. Won, K. Lee, C.-H. Lee, H.H. Lee, I.-C. Kim, J.-m. Lee, Graphene
2
oxide nanoplatelets composite membrane with hydrophilic and antifouling properties for
3
wastewater treatment, J. Membr. Sci., 448 (2013) 223-230.
4
[30] R. Yang, H. Jang, R. Stocker, K.K. Gleason, Synergistic Prevention of Biofouling in Seawater
5
Desalination by Zwitterionic Surfaces and Low‐Level Chlorination, Adv. Mater., 26 (2014) 1711-
6
1718.
7
[31] J. Meng, Z. Cao, L. Ni, Y. Zhang, X. Wang, X. Zhang, E. Liu, A novel salt-responsive TFC
8
RO membrane having superior antifouling and easy-cleaning properties, J. Membr. Sci., 461 (2014)
9
123-129.
10
[32] D. Saeki, T. Tanimoto, H. Matsuyama, Prevention of bacterial adhesion on polyamide reverse
11
osmosis membranes via electrostatic interactions using a cationic phosphorylcholine polymer
12
coating, Colloids Surf. Physicochem. Eng. Aspects, 443 (2014) 171-176.
13
[33] Y.-F. Yang, Y. Li, Q.-L. Li, L.-S. Wan, Z.-K. Xu, Surface hydrophilization of microporous
14
polypropylene membrane by grafting zwitterionic polymer for anti-biofouling, J. Membr. Sci., 362
15
(2010) 255-264.
16
[34] M.L.M. Tirado, M. Bass, M. Piatkovsky, M. Ulbricht, M. Herzberg, V. Freger, Assessing
17
biofouling resistance of a polyamide reverse osmosis membrane surface-modified with a
18
zwitterionic polymer, J. Membr. Sci., 520 (2016) 490-498.
19
[35] H.Z. Shafi, Z. Khan, R. Yang, K.K. Gleason, Surface modification of reverse osmosis
20
membranes with zwitterionic coating for improved resistance to fouling, Desalination, 362 (2015)
21
93-103.
22
[36] P. Król, P. Chmielarz, Recent advances in ATRP methods in relation to the synthesis of
23
copolymer coating materials, Prog. Org. Coat., 77 (2014) 913-948.
24
[37] J. Ran, L. Wu, Z. Zhang, T. Xu, Atom transfer radical polymerization (ATRP): a versatile and
25
forceful tool for functional membranes, Prog. Polym. Sci., 39 (2014) 124-144.
26
[38] Y. Zhang, Z. Wang, W. Lin, H. Sun, L. Wu, S. Chen, A facile method for polyamide membrane
27
modification by poly (sulfobetaine methacrylate) to improve fouling resistance, J. Membr. Sci., 446
28
(2013) 164-170.
24
1
[39] C. Liu, J. Lee, J. Ma, M. Elimelech, Antifouling thin-film composite membranes by controlled
2
architecture of zwitterionic polymer brush layer, Environ. Sci. Technol., 51 (2017) 2161-2169.
3
[40] D. Saeki, S. Nagao, I. Sawada, Y. Ohmukai, T. Maruyama, H. Matsuyama, Development of
4
antibacterial polyamide reverse osmosis membrane modified with a covalently immobilized
5
enzyme, J. Membr. Sci., 428 (2013) 403-409.
6
[41] D. Saeki, H. Karkhanechi, H. Matsuura, H. Matsuyama, Effect of operating conditions on
7
biofouling in reverse osmosis membrane processes: Bacterial adhesion, biofilm formation, and
8
permeate flux decrease, Desalination, 378 (2016) 74-79.
9
[42] S. Rosenberger, U. Krüger, R. Witzig, W. Manz, U. Szewzyk, M. Kraume, Performance of a
10
bioreactor with submerged membranes for aerobic treatment of municipal waste water, Water Res.,
11
36 (2002) 413-420.
12
[43] M.-h. Huang, Y.-m. Li, G.-w. Gu, Chemical composition of organic matters in domestic
13
wastewater, Desalination, 262 (2010) 36-42.
14
[44] N. Kim, D.H. Shin, Y.T. Lee, Effect of silane coupling agents on the performance of RO
15
membranes, J. Membr. Sci., 300 (2007) 224-231.
16
[45] V. Freger, J. Gilron, S. Belfer, TFC polyamide membranes modified by grafting of hydrophilic
17
polymers: an FT-IR/AFM/TEM study, J. Membr. Sci., 209 (2002) 283-292.
18
[46] J. Huang, W. Xu, Zwitterionic monomer graft copolymerization onto polyurethane surface
19
through a PEG spacer, Appl. Surf. Sci., 256 (2010) 3921-3927.
20
[47] J. Zhang, J. Yuan, Y. Yuan, J. Shen, S. Lin, Chemical modification of cellulose membranes
21
with sulfo ammonium zwitterionic vinyl monomer to improve hemocompatibility, Colloids Surf. B.
22
Biointerfaces, 30 (2003) 249-257.
23
[48] J.B. Schlenoff, Zwitteration: coating surfaces with zwitterionic functionality to reduce
24
nonspecific adsorption, Langmuir, 30 (2014) 9625-9636.
25
[49] D. Pavithra, M. Doble, Biofilm formation, bacterial adhesion and host response on polymeric
26
implants—issues and prevention, Biomedical Materials, 3 (2008) 034003.
25
1
[50] Q. Li, B. Zhou, Q.Y. Bi, X.L. Wang, Surface modification of PVDF membranes with
2
sulfobetaine polymers for a stably anti‐protein‐fouling performance, J. Appl. Polym. Sci., 125
3
(2012) 4015-4027.
4
[51] M. Herzberg, S. Kang, M. Elimelech, Role of extracellular polymeric substances (EPS) in
5
biofouling of reverse osmosis membranes, Environ. Sci. Technol., 43 (2009) 4393-4398.
6
[52] R. Kumar, A. Ismail, Fouling control on microfiltration/ultrafiltration membranes: Effects of
7
morphology, hydrophilicity, and charge, J. Appl. Polym. Sci., 132 (2015).
8
[53] G.-d. Kang, Y.-m. Cao, Development of antifouling reverse osmosis membranes for water
9
treatment: a review, Water Res., 46 (2012) 584-600.
10
[54] E.M. Vrijenhoek, S. Hong, M. Elimelech, Influence of membrane surface properties on initial
11
rate of colloidal fouling of reverse osmosis and nanofiltration membranes, J. Membr. Sci., 188 (2001)
12
115-128.
13
[55] D.J. Irvine, A.-V.G. Ruzette, A.M. Mayes, L.G. Griffith, Nanoscale clustering of RGD peptides
14
at surfaces using comb polymers. 2. Surface segregation of comb polymers in polylactide,
15
Biomacromolecules, 2 (2001) 545-556.
26